Mutation scanning and direct DNA sequencing of all 50 exons of ABCR were completed for 150 families segregating recessive Stargardt disease (STGD1). ABCR variations were identified in 173 (57%) disease chromosomes, the majority of which represent missense amino acid substitutions. These ABCR variants were not found in 220 unaffected control individuals (440 chromosomes) but do cosegregate with the disease in these families with STGD1, and many occur in conserved functional domains. Missense amino acid substitutions located in the amino terminal one-third of the protein appear to be associated with earlier onset of the disease and may represent misfolding alleles. The two most common mutant alleles, G1961E and A1038V, each identified in 16 of 173 disease chromosomes, composed 18.5% of mutations identified. G1961E has been associated previously, at a statistically significant level in the heterozygous state, with age-related macular degeneration (AMD). Clinical evaluation of these 150 families with STGD1 revealed a high frequency of AMD in first- and second-degree relatives. These findings support the hypothesis that compound heterozygous ABCR mutations are responsible for STGD1 and that some heterozygous ABCR mutations may enhance susceptibility to AMD.

Finally, we examined whether the phenotype‐associated known/candidate pathogenic variants could explain the patient's disease, andif the MAF in population‐matched control data (8.3kJPN) was relatedto disease prevalence. Patients were classified as “Solved” if theirgenotype was consistent with their clinical phenotype. Patients wereclassified as “Partially solved” when a heterozygous known/candidatepathogenic variant was detected in a recessive allele, but without anadditional variant in trans. Patients were categorized as “Unsolved” iftheir genotypes exhibited either no candidate pathogenic variants ormultiple heterozygous pathogenic variants that did not explain thephenotype clearly. Variants annotated as causal for solved patients arelisted in Supporting Information: Table S2. Novel variants identified inthis study are listed in the second sheet of Table S2. SupportingInformation: Table S3 shows the phenotypes and genotypes of solvedpatients.2.4 | Statistical analysisBefore counting the allele frequency in our cohort, the list ofpatients was modified to contain only the proband to avoid theoverrepresentation of pedigrees with larger numbers of affectedindividuals. Inter‐pedigree comparisons of genetic diagnoses evaluat-ing proband only and proband with family members were performedby the chi‐square test. Enrichments of the pathogenic variants ingenetically solved and unsolved patients were compared by the one‐sided binominal test. Allele frequencies were compared with thehighest allele frequencies among the 8.3KJPN, HGVD, ExAC_EAS, andgnomAD_EAS databases. Statistical analyses were performed byR (ver. 4.0.3).2.5 | Detection of RP1:c.4052‐4053ins328(Alu insertion)Previously reported primers were used to amplify the expectedAlu‐inserted region (Nikopoulos et al., 2019). Genomic DNA wasamplified with Prime Star (TAKARA) following the manufacturer'sSUGA ET AL . | 310981004, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/humu.24492 by Mie University, Wiley Online Library on [07/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

Finally, we examined whether the phenotype‐associated known/candidate pathogenic variants could explain the patient's disease, andif the MAF in population‐matched control data (8.3kJPN) was relatedto disease prevalence. Patients were classified as “Solved” if theirgenotype was consistent with their clinical phenotype. Patients wereclassified as “Partially solved” when a heterozygous known/candidatepathogenic variant was detected in a recessive allele, but without anadditional variant in trans. Patients were categorized as “Unsolved” iftheir genotypes exhibited either no candidate pathogenic variants ormultiple heterozygous pathogenic variants that did not explain thephenotype clearly. Variants annotated as causal for solved patients arelisted in Supporting Information: Table S2. Novel variants identified inthis study are listed in the second sheet of Table S2. SupportingInformation: Table S3 shows the phenotypes and genotypes of solvedpatients.2.4 | Statistical analysisBefore counting the allele frequency in our cohort, the list ofpatients was modified to contain only the proband to avoid theoverrepresentation of pedigrees with larger numbers of affectedindividuals. Inter‐pedigree comparisons of genetic diagnoses evaluat-ing proband only and proband with family members were performedby the chi‐square test. Enrichments of the pathogenic variants ingenetically solved and unsolved patients were compared by the one‐sided binominal test. Allele frequencies were compared with thehighest allele frequencies among the 8.3KJPN, HGVD, ExAC_EAS, andgnomAD_EAS databases. Statistical analyses were performed byR (ver. 4.0.3).2.5 | Detection of RP1:c.4052‐4053ins328(Alu insertion)Previously reported primers were used to amplify the expectedAlu‐inserted region (Nikopoulos et al., 2019). Genomic DNA wasamplified with Prime Star (TAKARA) following the manufacturer'sSUGA ET AL . | 310981004, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/humu.24492 by Mie University, Wiley Online Library on [07/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

Finally, we examined whether the phenotype‐associated known/candidate pathogenic variants could explain the patient's disease, andif the MAF in population‐matched control data (8.3kJPN) was relatedto disease prevalence. Patients were classified as “Solved” if theirgenotype was consistent with their clinical phenotype. Patients wereclassified as “Partially solved” when a heterozygous known/candidatepathogenic variant was detected in a recessive allele, but without anadditional variant in trans. Patients were categorized as “Unsolved” iftheir genotypes exhibited either no candidate pathogenic variants ormultiple heterozygous pathogenic variants that did not explain thephenotype clearly. Variants annotated as causal for solved patients arelisted in Supporting Information: Table S2. Novel variants identified inthis study are listed in the second sheet of Table S2. SupportingInformation: Table S3 shows the phenotypes and genotypes of solvedpatients.